[KineticLensman]

Worst. I had a manager part of whose job was to review my docs (reports and bids) before they were released. Their approach to reviewing was to start at the beginning and work through a doc until they found a red flag. They then refused to go further until that issue was fixed, which significantly increased the length of the process and meant that there was invariably more stress to come on the next iteration. A lot of their review comments boiled down to "I don't like this do it again differently". And I was lucky, I was at least able to write decent English - one of their comments aimed at a co-worker was "This is the incompetent ramblings of a halfwit".

When I became a reviewer myself I resolved never to be like this. If I didn't like something, I would always indicate what change I was expecting for the next review to pass and also why the change was necessary - what specific problem was being addressed. If I didn't like something but couldn't say why, then I taught myself to not say anything at all, typically because it was a cosmetic difference in my preferences vs. the authors. I also realised that a really good technique was to have a quick chat with authors before they had done too much, especially to make sure we all agreed on the aim and purpose of the doc, and my review criteria - to avoid me introducing unexpected hurdles at the last minute. This became a good way of turning the review process into an iterative learning process for junior authors.